S

Screen and
d ICD in
i High
Hi h Rate
R t Wells
W ll
Terje Moen
Reslink
Objective

 Understand keyy p
parameters in screen erosion
 Design principles used to reduce erosion risk

2
Agenda

 Reservoir Uncertaintyy
 Screen Erosion
- By plugging
- By flow
 Flow scenarios
 LineSlot versus High Rate LineSlot
 ICD
Screen plugging and well cleanup

 Slot opening must be sized to allow flow back of mud and

filter cake
– Clean formation sand do not plug a screen
– Mixed with mud or similar, plugging may occur
 Shell reported screen failure in offshore Netherland gas well
i 2003
in
– PLT showed 95% of the flux from 3 m only
– S
Screen erosion resulted in massive sand production
– Too narrow slot opening was used
The Particle Size Distribution

100

 The formation sand alone do 80

not plug a screen 60

 The screen must see a kill 40

pill!!! 20

0
1000 100 10

100

80

60

40

20

0
1000 100 10
How to Plug a Screen?

 Experience from mud flow back

test
- Circulation of conditioned mud
- Small volume of sand present in
flow loop
 Sand trapped by screen
 Mud could not flow through screen
covered with sand
 Flow focused on non covered area
resulting in complete plugging
 In this case, mud was dissolved in
oil
Screen collapse test

 Kill pill
ill ffailed
il d
 Flow through screen
- High pressure
- High velocity
 Erosion!
Bridging of LCM
C
Critical fforce?
?
Failure of LCM bridge

Velosity 2p
v  125 m/s

E i !
Erosion!
Screen failure 1

 Screen used to prevent sand production in drill stem test

i cased
in dhhole
l
- Unconditioned mud accidentally was flowing through
screen
 After perforation, mud was 2500
mixed with formation sand
2400 16.3
2300

Pa]
Pressure [psia]
2200

Pressure [MP
2100 14.3
2000
1900
1800 12.3
1700
1600
1500 10.3
04.08.06 04.08.06 04.08.06 04.08.06 04.08.06
15:00:00 15:15:00 15:30:00 15:45:00 16:00:00
Time
Screen failure 2

 A mixture of formation sand, proppant and

LCM formed a very low permeability “filter
cake” on the outside of the metal mesh
media.
 The mesh collapsed and failed when
exposed to > 1000 psi pressure drop at the
screens during
d i iinitial
iti l unloading
l di operations.
ti
 Once the mesh collapsed and failed, the well
began to flow 6% sand, further damaging
the screen from erosion.
erosion
Screen failure 3

 Mesh screen installed in drill in

fluid
 Filter cake shaved off and
trapped inside the screen
 No flow to surface
 In this case, screen was not
exposed
p to critical p
pressure
 Operator has stopped using
mesh type screen
Erosion Testing
g
 Screen sample
• Area
 Fluid
• Sand concentration
• Flow rate
 Result
• Mass loss

13
 Erosion Theory and Test Results

 Erosion rate is a function of velocity (u),

sand mass flow (m), impingement angle (α),
material constants and density (K, n and ρ)
(from Huser & Kvernevold, DnV)
 Erosion test results
• IRIS ((research
h iinstitute,
i S
Stavanger):) RF
RF-98/031
98/031
• Southwest Research Institute: Reslink and Slb tests
 K and n can be considered constants for
common screen materials 1.E-04
• K≈4x 10-7
1.E-05

ss / Mass
• n ≈ 2.5

asives
1.E-06
 Peak velocityy is a function of Mass los
abra 1.E-07
screen and ICD design
 10 times velocity increase 1.E-08
0.1 1 10
Velosity [m/s]
increases erosion ~ 300 times!

14
Erosion Rate 1E+04

 Assumptions: 1E+03

Erosion ratee [µm/year]

• Sand is continuously produced and 1E+02

disappears
pp through
g the screen 1E+01

• Sand loading: 4 x 10-6 kg/sm3 (0.25 lb/MMscf ) 1E+00

Average
• Impingement angle function F(α)=1 1E-01
10 x
• Formation volume factor B=0.0042 1E-02

 The local slot geometry increases 1E 03

1E-03
0.1 1 10
localized erosion effect on slot opening Velocity [m/s]
by a factor of 5 – 10 times
 0.3
0 3 m/s
/ should
h ld be
b considered
id d a sound
d ddesign
i velocity
l it
• Relates to the narrow tolerances of the slot opening
• Includes a reasonable safety margin

 Management of localized peak velocities becomes the

logic way of preventing screen erosion

15
Particle tracking to illustrate risk of erosion

 The flow starts converging

t
towards
d the
th slots
l t very close
l tto
the surface
 Particles approaching the
screen surface will not change
direction or velocity
 Any particle will hit screen
surface randomly
 Velocity distribution
determined by screen design
Inflow distribution and screen design

 No or pore drainage layer will give a near spherical flow approaching

the screen surface
- High perforation density is required to avoid high peak velocity
over perforations
 Good drainage area and slot opening area sized to drainage area
gives a more even flux towards the screen surface
- A normal perforation pattern that do not weaken the pipe can be
used
Wire wrap versus mesh

 Simulation conditions
- Equal perforation density
- Equal rate
 LineSlot
- High
g axial flow channel and limited area
to flow provides an even flux towards the
screen surface
- Particles are impinging the screen
surface more evenly
 Premium sintered mesh
- Shroud and limited drainage are focuses
the flow towards the base pipe
perforation
- Higher perforation density required and
also being used
Velocity plot in GP/collapsed annulus case

 The pack around the screen will distribute the flow and eliminate the erosion risk
 This provides a perfect pack of the whole interval
Rock fractures & peak velocities

 Small fractures can cause huge velocities.

 A 30mm thick fracture can cause peak velocity >100x
average inflow
i fl velocity.
l i
 Model illustrated is
5m radius axi-symmetrical 10m radius of rock
into reservoir (1
D
Darcy))
fracture into the well Velocity at wellbore
- A short fracture (<1m) will 12m section (0.02 m/s)

cause almost the same of well

Wellbore
peak in velocity; turbulent flow Peak in V (2.8
(2 8 m/s)

Driving pressure
in fracture will constrain corresponding to fracture
additional fracture flow

30mm thick
D

fracture

Radial Darcy flow

Annular Flow – Screen Only Completion

 Two flow paths

– Equal pressure drop
– Different cross sections determines flow distribution between
annulus and pipe
pp
– Can be calculated using basic analytical equations

 All annular flow enters at heel of last joint

– Potential for erosion
Annular flow - hotspotting

 Flow enters compartment toe in tubing

basepipe flow

annular flow
Annular flow - hotspotting

 Flow enters compartment toe in tubing

 Flow distributes between tubing and annulus

basepipe flow

annular flow
Annular flow - hotspotting

 Flow enters compartment toe in tubing

 Flow distributes between tubing and annulus
 Parallel tubing/ annular flow

basepipe flow

annular flow
Annular flow - hotspotting

 Flow enters compartment toe in tubing

 Flow distributes between tubing and annulus
 Parallel tubing/ annular flow
 Flow re-enters tubing
g at compartment
p heel

basepipe flow

annular flow
Annular flow - hotspotting

 Flow enters compartment toe in tubing

 Flow distributes between tubing and annulus
 Parallel tubing/ annular flow
 Flow re-enters tubing
g at compartment
p heel
 Potential erosion hotspot - high velocity

basepipe flow

annular flow
CFD Modelling of Peak Velocity
Perspective view

Heel
Toe
 Standard LineSlot screen design
 High rate 5.6x106 sm3/d (200 MMscfd): 0.65 m/s
- Down
D h
hole
l rate:
t 23 000 rm3/d (B
(Bg = 0.00402)
0 00402)
 Low rate 2.8x106 sm3/d (100 MMscfd): 0.33 m/s
- Down hole rate: 11 600 rm3/d
 High rate and 10% radial contribution: 0.96 m/s

27
LineSlot versus HR LineSlot

2% 2%
1%
1%
0%
-1% 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 0%
-2% 1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
-1%
-3%
-4% -2%
-5% -2.2 %
-3%
-6%
-7% -4%
-8% Peak rate reduced with a factor 3.6
-9%
Total rate can be increased acoringly

28
Pressure Profiles
 Introduction of controlled pressure drop in radial direction
• This is the principle behind conventional ICD completions
 Gas ICD
• Pressure drop: 0.1 – 1 MPa (15 – 150 psi)
• ICD has a maximum capacity
per joint (ICD housing) 20000
18000
 Gas screen 2.5
16000
• Pressure drop: ~ 5 kPa (0.7 psi)
14000 2
• The flow is forced to enter

Pressure [[psi]
Pressure [[Pa]
12000
a larger surface area 1.5
10000
• The flow capacity increases 8000
with improved distribution Tubing Pressure Ø10mm 1
6000
Annulus Pressure Ø10mm
• 5mm perforations can handle nearly 4000
T bi P
Tubing Pressure Ø5
Ø5mm 05
0.5
4 times the rate compared to 10mm perf.
2000
Annulus Pressure Ø5mm
• Screen design defines max 0 0
flow capacity of the well 0 2 4 6 8 10 12
Pos along screen joint [m]

29
ICD Simulations

 Out and in effect is solved by using ICD

 Screen next to ICD housing has a maximum flow capacity

- Effect of drainage layer
- Rate effect

 ICD interaction with reservoir

Effect of Drainage Layer

2
Relaative peak vvelocity

1.5

0.5

0
0 5 10 15 20
Drainage layer height [mm]
Peak Velocity as Function of Down Hole Rate
pr ICD Housing
[rB/d]
0 2000 4000 6000 8000 10000 12000 14000
0.7
0.6 2
05
0.5
1.5
0.4
[m/s]

[ft/s]
0.3 1
0.2
0.5
0.1
0 0
0 500 1000 1500 2000 2500
[rm3/d]
Simplified Reservoir Simulations

 When assuming 100 MMscfd

(11383 rm3/d),
/d) maximum
down hole rate is after 11
years (14093 rm3/d)
 Simplified 3 zone model
- 10 mD (13%)
- 40 mD (73%)
- 100 md (13%)
 8 x 4 mm nozzles and 15
j i t
joints
- Peak rate below critical
- Regulation too hard
Simplified Reservoir Simulations continued

 8 x 4 mm nozzles and 25
joints
 Startup rate 200 MMscfd
(22766 rm3/d)
- Below critical rate
 After 11 years (14093 rm3/d)
- Well below critical rate
- Regulation is reasonable
Robustness
Relative peak velocity
 The ability to handle unforeseen localized 1
high velocities 0.8
• Fractured reservoir 0.6
• Local high permeability 0.4
0.2
• Partly covered screen
0
This is not very likely as free particles will be
carried by the flow towards the unpacked
section
 HR LineSlot screen forces the flow to enter
0.6 1.8
over a larger area
0.5 1.6
 ICD provides a controlled pressure drop in the
range of the Darcy pressure drop through the 0.4 1.4

Relattive flux
e [MPa]
reservoir 0.3 1.2
• At a 1:1 regulation,
g , the maximum flow at an ICD can

Pressure
02
0.2 1
theoretically be:
0.1 0.8
qMAX  2 p ICD _ AVG  1.4q AVG 0 0.6
0 500
L
Length
th ffrom h
heell [[m]]

35
Summary

 Commonlyy used critical velocityy 0.3 m/s

 Screen based completion has an upper limit on total
production
- Standard LineSlot: ~10 000 rm3/d
- LineSlot-Gas: ~36 000 rm3/d
- Above numbers must be redused based on radial
contribution
 ResFlow completion
p has a maximum rate p
pr ICD
housing
- ~1000 rm3/d/ICD

36
37
38